THE JOURNAL OF COMPARATIVE NEUROLOGY 484:191–205 (2005) Projections from the Spinal Trigeminal Nucleus to the Cochlear Nucleus in the Rat CHARLES-ANDRE´ HAENGGELI,1 TAN PONGSTAPORN,1 JOHN R. DOUCET,1 AND DAVID K. RYUGO1,2* 1Department of Otolaryngology–Head and Neck Surgery, Center for Hearing Sciences, Johns Hopkins University School of Medicine, Baltimore, Maryland 21205 2Department of Neuroscience, Center for Hearing Sciences, Johns Hopkins University School of Medicine, Baltimore, Maryland 21205 ABSTRACT The integration of information across sensory modalities enables sound to be processed in the context of position, movement, and object identity. Inputs to the granule cell domain (GCD) of the cochlear nucleus have been shown to arise from somatosensory brain stem structures, but the nature of the projection from the spinal trigeminal nucleus is unknown. In the present study, we labeled spinal trigeminal neurons projecting to the cochlear nucleus using the retrograde tracer, Fast Blue, and mapped their distribution. In a second set of experiments, we injected the anterograde tracer biotinylated dextran amine into the spinal trigeminal nucleus and studied the resulting anterograde projections with light and electron microscopy. Spinal trigeminal neurons were distributed primarily in pars caudalis and interpolaris and provided inputs to the cochlear nucleus. Their axons gave rise to small (1–3 ␮m in diameter) en passant swellings and terminal boutons in the GCD and deep layers of the dorsal cochlear nucleus. Less frequently, larger (3–15 ␮m in diameter) lobulated endings known as mossy fibers were distributed within the GCD. Ventrally placed injections had an additional projection into the anteroventral cochlear nucleus, whereas dorsally placed injec- tions had an additional projection into the posteroventral cochlear nucleus. All endings were filled with round synaptic vesicles and formed asymmetric specializations with postsynaptic targets, implying that they are excitatory in nature. The postsynaptic targets of these terminals included dendrites of granule cells. These projections provide a structural sub- strate for somatosensory information to influence auditory processing at the earliest level of the central auditory pathways. J. Comp. Neurol. 484:191–205, 2005. © 2005 Wiley-Liss, Inc. Indexing terms: granule cells; mossy fibers; polysensory convergence; synapses Hearing involves more than processing information The substrate for some of this integration is evident in about vibrations in air. As an animal moves its head or the cochlear nucleus. The cochlear nucleus receives input pinna, the acoustic cues used to locate sounds in the en- from a variety of nonauditory structures. The sources vironment are changing. Knowledge of head and pinna include the dorsal column nuclei (Itoh et al., 1987; Wein- position is necessary for the auditory system to interpret these changes. Vestibular and visual information can be used to determine relative motion between the organism Grant sponsor: National Institutes of Health / National Institute on and a sound source. Proprioceptive input might help the Deafness and Other Communication Disorders; Grant number: DC04395; auditory system to distinguish between environmental Grant sponsor: Swiss National Science Foundation. and self-generated sounds (e.g., vocalizations, chewing). *Correspondence to: David K. Ryugo, Johns Hopkins University School of Medicine, 510 Traylor Research Building, 720 Rutland Avenue, Balti- Identifying a sound also requires learning and memory. more, MD 21205. E-mail: [email protected] That is, sounds made by conspecifics will differ from those Received 6 August 2004; Revised 15 October 2004; Accepted 1 December of a predator, and they will evoke different affective 2004 states. In this context, we seek pathways in the brain that DOI 10.1002/cne.20466 might integrate these various functions. Published online in Wiley InterScience (www.interscience.wiley.com). © 2005 WILEY-LISS, INC. 192 C.-A. HAENGGELI ET AL. berg and Rustioni, 1987; Wright and Ryugo, 1996; Li and ysis of sound localization cues (Young et al., 1995; Kanold Mizuno, 1997a,b), Scarpa’s ganglion and the medial ves- and Young, 2001). To what extent the trigeminal complex tibular nucleus (Burian and Gstoettner, 1988; Kevetter fits into this scheme is unknown. The goal of this study and Perachio, 1989; Bukowska, 2002), the trigeminal gan- was to describe the source and target(s) of the trigeminal glion (Shore et al., 2000) and nuclei (Itoh et al., 1987; pathway to the CN. Our working hypothesis is that the Wright and Ryugo, 1996; Li and Mizuno, 1997a,b; Wolff identification of these cells and circuits will help us un- and Ku¨ nzle, 1997), and the pontine nuclei (Ohlrogge et al., derstand the role of each pathway for sound perception. 2001). These projections emphasize the idea that the We used retrograde tracing methods to map the distri- meaning of a sound depends on more than its physical bution of cells in the spinal trigeminal nucleus that project characteristics. to the CN. We found that the predominant source for this In the dorsal cochlear nucleus (DCN) of the cat, the pathway originates in the caudal divisions of the nucleus. response properties of projection neurons (e.g., pyramidal In a second set of experiments, we injected anterograde and giant cells) are modulated by manipulating the pinna tracers in these caudal regions and examined the distri- or electrically stimulating the nerve carrying propriocep- bution and type of labeled terminals produced in the CN. tive afferents from areas that include the pinna (Saade´et Terminals were primarily observed in the GCD but many al., 1989; Young et al., 1995; Kanold and Young, 2001). were also located in the central core of the DCN and the These results indicate that somatosensory activity influ- ventral CN (VCN). The ultrastructural features of some ences auditory neurons but it is not entirely known which large terminals in the GCD defined them as mossy fibers. circuits of the somatosensory system convey the informa- They resemble those that arise from the cuneate nucleus tion. Some neurons in the primary relay nuclei of the and appear to target the dendritic claws of granule cells. somatosensory system send their axons into the cochlear nucleus (e.g., cat, Itoh et al., 1987; rat, Weinberg and Rustioni, 1987; Wright and Ryugo, 1996; Li and Mizuno, MATERIALS AND METHODS 1997; hedgehog, Wolff and Ku¨ nzle, 1997) but there are differences in the details of the observations. In the cat, Animals and animal preparation brainstem somatosensory projections were restricted ex- A total of 21 male Sprague-Dawley rats, weighing 250– clusively to the granule cell domain (GCD, Itoh et al., 350 g, were used in this study. Animals were deeply anes- 1987), whereas the projection in the rat was “almost com- thetized with an intraperitoneal injection of sodium pen- pletely confined” to the DCN, to the superficial granular tobarbital (45 mg/kg) and given an intramuscular layer of the ventral cochlear nucleus, and to the subpe- injection of 0.1 cc of atropine to reduce secretions. Lido- duncular corner of granule cells (Weinberg and Rustioni, caine was infiltrated around the ear canals and the sur- 1987). In contrast, more recent reports in rat and hedge- gical areas. When the animal was areflexic to a paw pinch hog report somatosensory projections throughout the GCD and had no corneal reflex, surgery was initiated. All pro- and the deep layers of the DCN (Wright and Ryugo, 1996; cedures were in accordance with established NIH guide- Wolff and Ku¨ nzle, 1997). The projection in the hedgehog, lines and approved by the Animal Care and Use Commit- however, is sparse compared to that observed in the rat. tee of the Johns Hopkins School of Medicine. Curiously, light touch to the head and neck and movement Retrograde labeling. We used retrograde tracers to of the vibrissae, all mediated through the trigeminal label cells in the trigeminal nuclei that project to the CN nerve, were ineffective in evoking responses in DCN neu- in the rat. The source of this pathway is located primarily rons in the cat (Kanold and Young, 2001). These varia- in the caudal regions of the spinal trigeminal nuclei (Itoh tions across species and/or methods merit further investi- et al., 1987; Wolff and Ku¨ nzle, 1997). However, neurons in gation if we are to understand the mechanisms of rostral portions of these nuclei (e.g., pars oralis) project to polysensory integration. the CN in the hedgehog tenrec (Wolff and Ku¨ nzle, 1997), Resolving these details is important because trigeminal whereas they may not in the cat (Itoh et al., 1987). Our terminals in the central core of the cochlear nucleus imply goal was to examine this issue in the rat and also provide a direct influence on the projection cells giving rise to the a target for the forthcoming anterograde studies. In these ascending auditory pathways. In contrast, trigeminal in- retrograde experiments, tracer was injected in the granule put confined to the GCD suggests an indirect effect on cell lamina (GCL) situated between the DCN and VCN auditory processing because the GCD is composed of in- because the GCD receives most of the input from trigem- terneurons (Mugnaini et al., 1980a; Wright and Ryugo, inal cells (Itoh et al., 1987; Wolff and Ku¨ nzle, 1997). 1996). We were also interested in the synaptic terminals A skin incision was performed to expose the occipital formed by this pathway in the GCD. The GCD contains a cranium. The posterior calvarium was drilled away in particular type of synaptic ending called a mossy fiber order to expose the underlying cerebellum. The dura was (McDonald and Rasmussen, 1971; Mugnaini et al., 1980b; cut and cerebellar tissue was aspirated in order to reveal Wright and Ryugo, 1996; Weedman et al., 1996). Mossy the dorsal and lateral aspects of the left cochlear nucleus. fibers are distinguished by their large size, irregular Using an operating microscope, a glass micropipette (40 shapes, and numerous synapses formed with microneu- ␮m tip ID) filled with a 3% aqueous solution of Fast Blue rons in the GCD (Mugnaini et al., 1980b; Hutson and (FB, Sigma Chemical, St.